Measuring the Unequal Gains from Trade

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

This project will use new methods to answer questions about how international trade affects income inequality. Economists have already examined how international trade affects the distribution of earnings across individuals; in other words, how trade affects wages and salaries. However, trade also affects the prices people pay for everything from food to electronics. These price differences may be a second 'channel' for how trade affects inequality. For example, if lowering trade barriers reduce the price of food, low-income consumers, who typically spend a lot on food, may be better off relative to high-income consumers. Conversely, if international trade lowers the price of high-end electronics, this may benefit the rich relatively more than the lower-income consumers. While economists understand that this expenditure channel is possible, measuring the effects of the channel is important in its own right. This project develops a methodology to measure the distribution of welfare changes through this expenditure channel. Some of the questions the PIs seek to answer are: How important are the distributional effects of trade through this expenditure channel? Does trade tend to benefit high- or low-income consumers within a country? How do these effects vary across countries depending on their level of development? The answers to these questions will advance science. Insights from the work may help policy-makers better understand the effects of trade agreements, which could promote the national prosperity. This project develops a methodology to measure the distribution of welfare gains from trade across individuals through changes in the cost of the consumer basket. The methodology is based on aggregate statistics and is therefore broadly applicable across countries and time periods. The approach uses the Almost Ideal Demand System (AIDS). Using properties of this demand system, we demonstrate that the first-order approximation to the welfare change (i.e., the compensating variation) of a given consumer caused by an arbitrary change in prices can be recovered using demand parameters and aggregate statistics. Embedding this demand-side result in a standard quantitative model of international trade allows us to use aggregate trade shares and demand-side parameters (the income and price elasticity of imports by source country and sector) to characterize the welfare change at each income level in response to a trade shock. In turn, the demand side parameters can be estimated using aggregate trade and production data from a non-homothetic gravity equation generated by the model. Preliminary calculations suggest that the welfare effects of trade through the expenditure channel trade typically favors the poor, who concentrate spending in more traded sectors. The second part of this project applies the insights of our methodology to a unique dataset that tracks barcode-level purchases by households in Argentina. We use the framework to explore how a large devaluation of the Argentine Peso differentially affected households across the expenditure distribution. We also document unequal welfare changes through the expenditure channel during periods of economic boom and contraction. To our knowledge, this would be the first project to track extremely detailed consumption data across households during a devaluation episode and over the business cycle.